I have a data : 24.4±0.015
I want to find out the maximum without having to write a code.
How is it possible?

For example, the so-
http://smath.info/live/?file=3857

Hey, this would be a very elegant way for the analysis of tolerance chains.

But there would be a Problem:

How can you enter a value of e.g. 25 +0,1/-0,2 or 25 +0/-0,2 ??

Any ideas?

Hi,

Wrote

How can you enter a value of e.g. 25 +0,1/-0,2 or 25 +0/-0,2 ??

It's not difficult, just use the multiple values item (type "sys(" or click on the curly bracket in the "functions" right panel)

[LIVE width=259 height=195]http://smath.info/live/?file=3861[/LIVE]


regards,

w3b5urf3r

Thanks w3b5urf3r,

but then I have the same question as in the beginning of this thread: how do I get the min / max-value of this result?

example

(btw how did you enter the preview of the live file in your post?)

HI

Wrote

Thanks w3b5urf3r,

but then I have the same question as in the beginning of this thread: how do I get the min / max-value of this result?

example

(btw how did you enter the preview of the live file in your post?)

For the max() and min() values use the snippet sys2mat() in the attachment

To show the preview click on "Get link..." in the bottom-right corner of Live SMath and copy&paste in your post the 2nd link


best regards,

w3b5urf3r

---

sys2mat_example.sm (11 KiB) downloaded 67 time(s).

Great - but one problem:

I get an error message "mat()-function is not defined" on every term below the line tc:= sys2mat(tc)

I use Smath 0.95 build 4594 on linux (same also on version 0.94 on linux).

Hello,

Here is a screenshot from Ubuntu 12.04
No errors reported.

Regards,
Radovan

---

Hmmm - this looks like it's my computer (running on the older ubuntu 8.04 (server version)) that have this problem. Update is scheduled, let's see how it changes.

Another question: Does SMath have something like a median-Funktion (= find the value between two given numbers)? Because the usual production process tries to work in the middle of the tolerance field this value could also be interesting to know in a tolerance chain.

Btw, why isn't it possible to transcode a system into a matrix with a build-in function? Or enhance the min/max/lenght/... function to work also with a system and not only with a matrix/vector? Is this not a common problem?

Hello,

Wrote

Hmmm - this looks like it's my computer (running on the older ubuntu 8.04 (server version)) that have this problem. Update is scheduled, let's see how it changes.

Try this SMath DEB packages. During the installation, missing dependencies will be reported, if any.

Regards,
Radovan

Hi

Wrote

Another question: Does SMath have something like a median-Funktion (= find the value between two given numbers)? Because the usual production process tries to work in the middle of the tolerance field this value could also be interesting to know in a tolerance chain.

You can try with linterp() or building a function like [MATH]med(A;A+/2[/MATH]

Wrote

Btw, why isn't it possible to transcode a system into a matrix with a build-in function? Or enhance the min/max/lenght/... function to work also with a system and not only with a matrix/vector? Is this not a common problem?

I don't know, it's not yet implemented but SMath is in continuous development and something can escape... I'm sure that in a future version there will be some improvements, specially if you report your Feature Requests... f.e. an improvement of max() and min() is requested from several users (max and min), and the support of systems is something that should be included but it's not reported


regards,

w3b5urf3r

Hello,

Why I am getting different answers?
(http://www.yourfilelink.com/get.php?fid=816752&dv=1)




Thanks,
Basile

Wrote

Hello,

Why I am getting different answers?
(http://www.yourfilelink.com/get.php?fid=816752&dv=1)




Thanks,
Basile

yourfilelink.com report: "Incorrect linking code. Please try again."

can't you attach the file on your post?

w3b5urf3r

Here should be...

---

Wrote

Here should be...

Hi,

it's the optimization in the definition of [MATH]R[/MATH], [MATH]R,eq[/MATH] and [MATH]R,e[/MATH] (see optimization of coloured equations in the attachment)


regards,

w3b5urf3r

---

sys.sm (13 KiB) downloaded 37 time(s).

Hello,

Pay attention how sys() is working. In this simple example, every element from both sys() are combined with each other:

[MATH=eng]sys(1,20,2,1)+sys(50,300,2,1)=sys(51,70,301,320,4,1)[/MATH]

Sometimes it behaves differently, it depends on optimization applied.

Optimization numeric:
[MATH=eng]sys(1,20,2,1)+sys(1,20,2,1)=sys(2,21,21,40,4,1)[/MATH]

Optimization symbolic:
[MATH=eng]sys(1,20,2,1)+sys(1,20,2,1)=2*sys(1,20,2,1)[/MATH]

Actually: [MATH=eng]2*sys(1,20,2,1)=sys(2,40,2,1)[/MATH]

To be honest, because sys() in the calculations can sometimes be quite confusing, my advice is to avoid it in calculations. Actually, I did not figure out yet where the sys() environment is the most useful for using - excluding plots.

Regards,
Radovan

Thank you both for your replies.
But how somebody will know a priori to leave the default or change the optimization?

Thanks,
Basile

Hello w3b5urf3r,

I found the reason why your snippets don't work on my machine. Change the decimal symbol to comma and the arguments separator to semicolon and you get the same errors. With decimal symbol "period" and argument separator "comma" everything is fine.

Haven't tested if this is a bug or a feature...

Because Andrey is improving the min/max-function in near future these snippets won't be needed any more (probably/hopefully)

greetings
jabadu

Wrote

Hello w3b5urf3r,

I found the reason why your snippets don't work on my machine. Change the decimal symbol to comma and the arguments separator to semicolon and you get the same errors. With decimal symbol "period" and argument separator "comma" everything is fine.

Haven't tested if this is a bug or a feature...

Because Andrey is improving the min/max-function in near future these snippets won't be needed any more (probably/hopefully)

greetings
jabadu

Thank you very much...


seem it's an issue (BUG?) of the built-in functions [MATH]str2num(#)[/MATH] and [MATH]num2str(#)[/MATH] that have different behavior changing interface settings (see the attachments)


(probably the wrong behavior is in [MATH]num2str(#)[/MATH] but I'm not sure)


regards,

w3b5urf3r

---

num2str_str2num_bug.sm (5 KiB) downloaded 38 time(s).

Thank you! Bug of str2num function fixed.

Regards.